1. HYPOTHESIS 1592U89, an agent for treatment of HIV infection in children can be used safely in children as an alternative to currently employed agents, or in combination therapy regimens. 2. SPECIFIC AIMS A. To assess the steady-state pharmacokinetic features, tolerance, and safety of orally administered 1592U89, given alone or in combination with other antiretroviral medications, in HIV-infected infants and children. B. To establish doses of 1592U89 appropriate for future Phase II/III clinical trials. C. To examine potential age-associated differences in the pharmacokinetics, tolerance, and safety of 1592U89, used alone or in combination with other antiretroviral agents. D. To generate preliminary information on the effects of 1592U89, used alone or in combination with other antiretroviral medications, on virus load and surrogate immunologic markers. 3. BACKGROUND On the basis of preclinical and clinical studies 1592U89 appears to be a promising agent for treatment of HIV infection in children, either as an alternative to currently employed agents, or in combination therapy regimens. The drug 1592U89 succinate is a carbocyclic 2',3'-ene nucleoside analog with in vitro activity against HIV similar to that of zidovudine (ZDV). In vitro selection for resistance to 1592U89 is slow to occur, and laboratory and clinical isolates of HIV with decreased susceptibility to ZDV are not cross- resistant to 1592U89. The antiviral activity of 1592U89 in combination with other antiviral agents has been assessed in MT-4 cells infected with HIV-1 strain IIIB. 1592U89 was synergistic with ZDV and 3TC. The combination of 1592U89 and ddI showed slight synergy in some experiments, while the combination of 1592U89 and d4T showed variable results ranging from slight synergy to slight antagonism. Serious adverse clinical events or laboratory abnormalities were not observed among 18 HIV-infected adults receiving 1592U89 in single oral doses ranging from 100 mg to 1200 mg. A multiple dose Phase I study of 1592U89 (200, 400 or 600 mg orally every eight hours) is ongoing. Preliminary results indicate that the drug is well tolerated and safe. Subjects entered the study with a mean CD4+ lymphocyte count of 342/ L; the mean changes after four and 12 weeks of treatment were +80/ L and +120/ L, respectively. The mean plasma HIV RNA copy number per mL at entry was 5.2 log, with decreases of 1.5 log and >2 log at four and eight weeks, respectively. Eighteen infants and children aged six months to 13 years have received 1592U89 (4 or 8 mg/kg orally) in a Phase I single dose study. Serious adverse events have not been observed; however, two patients developed asymptomatic elevation of alkaline phosphatase.